High-Gain Observer-Based Identification Scheme for Estimation of Physical Parameters of Synchronous Generators

This paper deals with estimation of states and physical parameters of synchronous generators which is of great significance in power system analysis as well as control. To handle the difficulties associated with the existence of unknown nonlinearities in generator dynamics, the system dynamics is firstly transformed into a canonical form using a change of variables leading to an equivalent system. Then, a robust observation scheme is proposed using Phasor Measurements Units' (PMUs') data along with a combination of the well-known Genetic Algorithm (GA) and a modified version of high -gain observers. The physical parameters of synchronous generator are identified by decomposing the nonlinear function of the system dynamics into a regression model. This decomposition enables us to identify the unknown parameters accurately by using the estimated state variables and Recursive Least Square (RLS) technique. Finally, the proposed identification scheme is compared with the well-known Iterative Extended Kalman Filter (IEKF) technique throughout simulations. The obtained results approve the theoretical discussions and demonstrate the superiority and feasibility of the proposed identification methodology.